The Canadiens dominated the dangerous areas—six high-danger chances to Buffalo's two—but lost 5–3. The underlying metrics called the game close to even: expected goals finished 2.99–3.23, a slight Montreal edge. The Sabres' conversion rate turned that into a decisive win.

Buffalo's finishing was elite. The Sabres overperformed their expected goals by 2.01 while generating just 0.053 expected goals per shot compared to Montreal's 0.075—lower volume, higher conversion. Montreal crashed the net for four rebound chances (converting two) and generated superior looks, but the Sabres struck with lethal efficiency on the rush, scoring five goals from twenty-six rush chances.

Goaltending diverged sharply. J. Fowler (MTL) surrendered 1.0 goals more than expected on his workload—a weak performance. C. Ellis (BUF) posted a 0.870 save percentage and controlled the game when it mattered.

Special teams tilted toward Buffalo. The Sabres power play converted twice on ten shots (0.56 xG). Montreal matched the goal total—2 PPG on 6 PP shots (0.63 xG)—but trailed and couldn't recover. The turning point came on a Tage Thompson wrist shot from the high slot at 18:20 of the third, cementing Buffalo's control.

Buffalo outshot Montreal 27–23 on goal (54 percent of attempts) but that doesn't explain the margin. The Canadiens created the better chances; the Sabres finished theirs.

The pregame model offers no favors here: Montreal-Buffalo projects to a 2.5–2.7 expected goal spread with Buffalo holding 54% win probability. That's a pick'em.

The advanced stats explain why. Montreal owns 48.0% 5v5 expected goals and 49.0% Corsi; Buffalo counters with 50.0% xG% and the same 49.0% Corsi%. No team has established clear territorial dominance.

Both offenses, though, are running hotter than the underlying metrics suggest. Montreal's PDO sits at 1.018; Buffalo's at 1.020. When shooting percentage and save percentage both exceed their natural averages, regression is the house edge. Expect some evening out here.

S. Montembeault will draw scrutiny on the back-to-back: watch for Montreal to pivot to a backup, or for Montembeault to show fatigue in the third.

For Montreal, Cole Caufield commands the most attention. His 0.97 expected goals per 60 minutes would rank among league leaders; he's converted that pace into 33 goals with just 1 high-danger goal, suggesting he's scoring on lower-quality chances at a 20.0% clip—which is sustainable but unsustainable in a regression season. Oliver Kapanen trails in xG/60 (0.83) but has four high-danger goals to Caufield's one, a sign his offense lives closer to where the danger actually is.

Buffalo's scoring mirrors Montreal's imbalance. Jason Zucker generates 1.07 xG/60 and has buried 6 of his 14 goals from high-danger spots, meaning he's getting prime real estate. Josh Doan's 1.01 xG/60 has yielded 15 goals but just 2 high-danger tallies—a shooting rate unsupported by high-danger creation.

Defensively, Montreal's Mike Matheson and Buffalo's Jack Quinn are the identified weak links in transition. Whichever team can isolate and exploit their opposing weak link will inch toward the 54% figure in their favor.

This is the narrowest of margins: two evenly matched teams, both running hot, both vulnerable to regression.